A new two-stage sparse representation for fast multi-target location in bistatic MIMO radar

The multi-target location in bistatic MIMO radar under the colored noise is studied in this paper, and a two-stage sparse representation (TSR) method for joint DOD and DOA estimation is proposed. This method obtains the signal subspace by eigenvalue decomposition of the fourth-order cumulants (FOC) matrix. Then, based on the redundant dictionary refined gradually, a sparse linear model is constructed and solved to obtain angle estimation by sparse recovery algorithm. The simulation results demonstrate that the proposed method suppresses the effect of the Gaussian colored noise, reduces the calculation of sparse recovery, and performs robustly even under low snapshots.